LA Arrest Visualization

Dependencies	Purposes
data.table	faster read large dataset
tidyverse	dplyr, tidyr, and ggplot
ggmap	maps API
maps	map API toolkits
mapdata	map data
ggrepel	manipulations of layers on maps
varhandle	unfactor function
gridExtra	show 2 maps side by side
options	number digits to display

library(data.table) 
library(tidyverse)  
library(ggmap)      
library(maps)       
library(mapdata)    
library(ggrepel)
library(ggpubr)
library(varhandle)
library(gridExtra)
options(digits=16, warn=-1)
ggmap::register_google(key='AIzaSyAyDbezM1mVF2yr5FnpX9bJ61zEVapyjqU')

1. Read the relevant dataset into a data frame
2. View the data frame

# Copy the dataset to a data frame
data_path = '~/Arrest-data-from-2010-to-present.csv'
la_arrests <- as.data.frame(fread(data_path))
View(la_arrests)

1. Select relevant variables in the dataset
2. Clean and wrangle data (variable names, unfilled values, latittue, longitude, date)
3. Factor variable $Charge_Group_Description
4. Drop messy variable $Location
5. Review the complete dataset

# Select relevant cols in the dataset
df <- select(la_arrests, `Arrest Date`, `Time`, `Age`, `Area Name`, `Sex Code`, 
             `Charge Group Description`, `Arrest Type Code`, `Location`)

# Rename the multi-word variables so that each of them does not have space
setnames(df, 
         old=c('Arrest Date', 'Area Name', 'Sex Code', 
               'Charge Group Description', 'Arrest Type Code'), 
         new=c('Arrest_Date', 'Area_Name', 'Sex_Code',
               'Charge_Group_Description', 'Arrest_Type_Code'))

# Extract latitude as double format from Location of the data frame
# Convert $Arrest_Date to Date format
df <- transform(df, 
                Latitude  = as.double(str_sub(word(Location, 2, 2), 2, -3)),
                Longitude = as.double(str_sub(word(Location, -1, -1), 2, -3)),
                Arrest_Date = as.Date(str_sub(df$Arrest_Date, 1, 10)))

# Fill emppty values in Charge_Group_Description variable with 'Unknown'
df$Charge_Group_Description[df$Charge_Group_Description==''] <- 'Unknown'

# Factor and sort $Charge_Group_Description
df <- within(df, Charge_Group_Description 
                 <- factor(Charge_Group_Description, 
                    levels=names(sort(table(Charge_Group_Description)))))

# Remove Location off the data frame
df <- select(df, -Location)

# Inpsect the data frame
View(df)
glimpse(df)

## Observations: 1,276,160
## Variables: 9
## $ Arrest_Date              <date> 2019-06-22, 2019-06-22, 2019-06-22, 20…
## $ Time                     <int> 1630, 1010, 400, 302, 1240, 800, 40, 83…
## $ Age                      <int> 44, 8, 31, 23, 28, 13, 31, 40, 20, 14, …
## $ Area_Name                <chr> "Pacific", "West Valley", "N Hollywood"…
## $ Sex_Code                 <chr> "M", "M", "F", "F", "M", "M", "M", "M",…
## $ Charge_Group_Description <fct> Miscellaneous Other Violations, Unknown…
## $ Arrest_Type_Code         <chr> "M", "O", "M", "M", "F", "D", "F", "M",…
## $ Latitude                 <dbl> 33.9920, 34.1687, 34.1649, 34.2692, 33.…
## $ Longitude                <dbl> -118.4201, -118.5579, -118.3965, -118.4…

• Plot a bar chart for $Charge_Group_Description in presenation of count and percentage in the same chart

p_colors <- c('#FC0BF0','#52FD4D','#F4B942','#350113','#F90359','#320EFC','#18B4B2')
p_colors <- c(rep(p_colors, 4))
ggplot(df, aes(x=Charge_Group_Description)) +
  geom_bar(color='black', fill=p_colors) +
  geom_bar(mapping=aes(x=Charge_Group_Description, y=..prop.., group=1),  
           stat='count', fill=p_colors) +
  geom_text(aes(label=paste(round(stat(..prop..)*100, 1),'%', sep=''), group=1), 
            stat='count', size=3.5, hjust=-0.07, color=p_colors, fontface='bold') +
  coord_flip() +
  labs(title='Los Angeles Drug Violations By Year\n(January, 2010 - June, 2019)', 
       caption='Source: https://data.lacity.org') +
  theme(axis.text.y=element_text(color=p_colors, size=12, face='bold'),
        axis.text.x=element_text(color='red', size=12, face='bold'),
        axis.title.x=element_blank(),
        axis.title.y=element_blank(),
        plot.title=element_text(color='#52FD4D', face='bold', hjust=0.5)) 

• Plot a bar chart for drug violations by year

drug_by_year <- data.frame(Year=integer(), Counts=double())
year_span <- 2010:2019
for (year in year_span) {
  cnts <- nrow(select(filter(df, Charge_Group_Description=='Narcotic Drug Laws', 
                                 as.numeric(format(Arrest_Date,'%Y'))==year)))
  drug_by_year[nrow(drug_by_year) + 1, ] = list(year, cnts)
}

ggplot(drug_by_year, aes(x=Year, y=Counts)) +
  geom_point(size=6, color='#52FD4D') +
  geom_segment(aes(x=Year, xend=Year, 
                   y=0, yend=Counts), color='#52FD4D') +
  scale_x_continuous(breaks=year_span) +
  geom_smooth(method='lm', se=FALSE, color='#FC0BF0') +
  labs(title='Los Angeles Drug Violations By Year\n(January, 2010 - June, 2019)', 
       caption='Source: https://data.lacity.org') +
  theme(axis.text.y=element_text(color='red', size=12, face='bold'),
        axis.text.x=element_text(color='red', size=12, face='bold'),
        axis.title.x=element_blank(),
        axis.title.y=element_blank(),
        plot.title=element_text(color='#52FD4D', face='bold', hjust=0.5)) 

• Function to plot a distribution map for drug violations

map_plot <- function(dataset, center) {
m <- ggmap(get_googlemap(center=c(lon=center[1], lat=center[2]), 
                         maptype='roadmap', scale=2, zoom=10))
m + 
  geom_point(aes(x=Longitude, y=Latitude), 
               data=dataset, color='#F90359', size=1) +
  stat_density2d(data=dataset, 
                 aes(x=Longitude, y=Latitude, alpha=..level..),
                 size=0.01, bins=8, geom='polygon') +
  scale_fill_gradient(low='#18B4B2', high='#F90359') +
  scale_alpha(range=c(0.5, 0.8), guide=FALSE)
}

• Plot a geographical map of drug violations:

1. Create a list of data frames which contains the terms ‘Narcotic Drug Laws’ and years of interest
2. Feed the each of data frames in the list in the plot function to draw its map

drugs <- list()
d_plot <- list()
for(year in year_span) {
  drugs[[year]] <- select(filter(df, Charge_Group_Description=='Narcotic Drug Laws', 
                          str_sub(Arrest_Date, 1, 4)==as.character(year)), 
                          `Latitude`, `Longitude`)
  d_plot[[year]] <- map_plot(drugs[[year]], c(-118.3533783, 34.0274463)) +
                          annotate('text', x=-118.1, y=34.35, 
                                   label=year, color='#FC0BF0', size=12)
}

## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx
## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx
## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx
## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx
## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx
## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx
## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx
## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx
## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx
## Source : https://maps.googleapis.com/maps/api/staticmap?center=34.027446,-118.353378&zoom=10&size=640x640&scale=2&maptype=roadmap&key=xxx